Protein Structure via Crystallisation

 Research based on protein crystallography has returned 11 Nobel Prizes to date.

Collective research concerning crystallography has resulted in 30 Nobel Prizes to date.

In order to obtain the atomic coordinates of a macromolecule using diffraction techniques, a pure sample of the protein must be generated; this sample must be subsequently coaxed into crystals which are then irradiated with X-rays, thus enabling determination of the structure.

Left to Right: Crystal, Diffraction, Density and Structure 

We can’t predict the mechanisms that lead to well diffracting crystal formation, therefore the current state of the art in the field relies on large numbers of very small (nano-scale) crystallisation experiments to allow us to navigate toward a successful condition.  Nuclear magnetic resonance spectroscopy (NMR) allows structural determination in solution, but has limitations regarding the size of the molecules that can be analysed. Watch Georgina Ferry give an excellent overview of the history of crystallography on the video below – thank you to Georgina and the Wellcome Trust for allowing us to share the video.

A (very) brief history

Biological macromolecules were successfully crystallised during the early 20th century, and X-rays were being used to determine inorganic crystal structures at around the same time.  The technologies soon combined, paving the way for the determination of the atomic coordinates of significant bio-macromolecules (Vitamin B12, myoglobin) – both recognised by Nobel Prizes.  Within the next decade, structures of lysozyme (the first enzyme structure) and insulin (the first human hormone structure) had also been solved.

Significant historical references

1955: Hodgkin et al solve the structure for Vitamin B12

  • The Crystal Structure of the Hexacarboxylic Acid derived from B12 and the Molecular Structure of the Vitamin. Nature. 176: 325-328.

1958: Perutz and Kendrew solve the structure for Whale Myoglobin

  • A Three-Dimensional Model of the Myoglobin Molecule Obtained by X-Ray Analysis. Nature. 181: 662-666.

1965: Phillips et al solve the structure of Lysozyme

  • Structure of Hen Egg-White Lysozyme: A Three-dimensional Fourier Synthesis at 2 Å Resolution. Nature. 206:757-761.

1969: Hodgkin et al solve the structure for Human Insulin

  • Structure of Rhombohedral 2 Zinc Insulin Crystals. Nature. 224: 491-495.

How do crystals form?

With few exceptions, bio-macromolecules do not form crystals in their native environment and altering that environment by changing the pH, adding some salt, changing the temperature, etc usually results in the formation of a precipitate (unfolded/aggregated protein).  The accepted pathway for growing crystals is to start with a well-folded protein and find a crystallant that creates a super-saturated state, and then very slowly decreases the solubility of protein over time; this is visually explained using the phase diagram below (the C3 uses the Vapour Diffusion approach).

Protein crystallisation phase diagram, we’re trying to push the protein gently into a state of supersaturation and then wait for a nucleation event.

Venn diagram illustrating the three primary variables, and some of the immediate sub-variables to consider. The protein quality should not be variable!

As it is hard to predict the conditions that will enable a protein crystal to form, many experiments are conducted to find the best (and worst) range of conditions (chemical and physical) that can be subsequently optimised to provide an environment in which crystals will grow – the Venn diagram above gives you a rough idea of the three major variables, which include almost infinite sub-variables.

To this end – given that the amount of protein sample is invariably the limiting factor – robotic technology and automated imaging are used to help identify the right condition. Sadly, the journey of structure determination is far from complete with the formation of a protein crystal; it must be well-ordered, single and ‘large’ (10’s of micron in each dimension) in order to create a high resolution X-ray diffraction pattern.

These shapes represent protein crystals, with each sub-unit representing an individual protein unit, from left to right; well ordered, poorly ordered and heterogeneous. The ordered shapes would create a high resolution diffraction pattern. As the shapes decrease in order, or as objects other than hexagons make their way into the structure (i.e. impurities, unfolded protein, etc), the quality of the resultant diffraction patterns will decrease.